By Rob Mitchum // August 8, 2017
For the future of the planet, there are few research subjects more important than the global supplies of food, water, and energy. To comprehensively study, understand, and inform policy around these complex systems, the next generation of researchers in the physical, social, and biological sciences will need fluency with data analysis methods that traverse traditional academic boundaries.
A new interdisciplinary University of Chicago curriculum will train graduate students from geosciences, economics, computer science, public policy, and other programs in computational and data science techniques critical for modern science. With a $3 million award from the National Science Foundation, the new research traineeship grant will combine expertise from across UChicago and Argonne National Laboratory in computing, statistics, social science, climate, and agriculture.
“This program will equip graduate students with the tools needed to advance the study of issues related to food, energy, and water,” said Elisabeth Moyer, associate professor of atmospheric science in the Department of the Geophysical Sciences at UChicago. “Our vision is to produce students who have the computational skills and breadth of knowledge, from social to physical sciences, needed to tackle these critical research subjects in all their complexity.”
As Earth’s population rises in the coming years, demand for food, energy, and water is expected to soar. The global scale and interdependence of these sectors — where increased agriculture decreases freshwater supplies, while both are affected by the environmental impact of accelerating energy production — necessitates research collaboration across fields. Improved data collection and modeling on these topics creates promising opportunities for understanding their complexities, but only if analyzed with the right computational methods.
The UChicago NRT program will produce students with a foundation in a discipline such as geosciences, economics, or public policy, as well as the computational skills and breadth of multidisciplinary knowledge to tackle complex questions in food, energy, and water. A three-year curriculum including bootcamps, retreats, new courses, and practicum projects will give each student experience working with data science methods and scientists from fields other than their own.
Additional training in scientific communication and professional development, as well as opportunities for international research experience with the Potsdam Institute for Climate Impact Research and the African Institute for Mathematical Sciences, will further prepare students for research careers in this area.
“We want to extend education outside the boundaries of traditional silo-ed disciplinary programs,” said Ryan Kellogg, professor at the Harris School of Public Policy. “This program will provide students with the computational skills needed to exploit the growing torrent of relevant data, and give them experience both interacting across disciplines and translating results to non-academic audiences.”
As part of the curriculum, students will receive introductions to computing in the social sciences and geosciences, spatial statistics and imagery analysis, geographic information systems (GIS), data science fundamentals, time series analysis, and environmental economics. A general course on the food/energy/water system, drawn from existing courses for interdisciplinary audiences taught by Moyer and Cristina Negri, Environmental Engineer at Argonne and Fellow of the Institute for Molecular Engineering, will be followed by a data analysis practicum where groups of students work with real data and organizations in government and industry.
The program will build upon successful UChicago training initiatives such as the Executive Program in Applied Data Analytics, the Computational Analysis and Public Policy curriculum at the Harris School of Public Policy, and the Data Science for Social Good Summer Fellowship.
Instruction and mentorship will be provided by several UChicago research groups, including the Center for Robust Decision-Making on Climate and Energy Policy (climate and agricultural modeling), Knowledge Lab (text mining), the Energy Policy Institute at UChicago (environmental and energy economics), the Center for Data Science and Public Policy (data analytics and project management), and the Center for Spatial Data Science (spatial analysis). High-performance computing resources and tutorials will be provided by the Research Computing Center.
“All across the University of Chicago campus, we have researchers applying innovative data science techniques to important questions in energy and the environment,” said Michael Franklin, Liew Family Chair of Computer Science. “With this new program, we can extend that strength to enrich our graduate education, making our campus a laboratory for training a new generation of interdisciplinary, computational-minded scientists.”
The award is one of 17 given out by the National Science Foundation last month as part of their NSF Research Traineeship (NRT) program. The foundation awarded a total of $51 million to “develop and implement bold, new, potentially transformative models for graduate education in science, technology, engineering and mathematics (STEM) fields.”
“Integration of research and education through interdisciplinary training will prepare a workforce that undertakes scientific challenges in innovative ways,” said Dean Evasius, director of the NSF Division of Graduate Education in a news release. “The NSF Research Traineeship awards will ensure that today’s graduate students are prepared to pursue cutting-edge research and solve the complex problems of tomorrow.”
In addition to Moyer, Kellogg, and Franklin, additional investigators on the award include Joshua Elliott, Computation Institute Fellow and research scientist, and Ian Foster, Arthur Holly Compton Distinguished Service Professor of Computer Science and Argonne Distinguished Fellow.
Image Caption: Median potential end-of-century renewable water abundance/deficiency (from Elliott et. al, PNAS, 2014)